About 15 pct of recent underground coal mine fires in the United States were caused by the spontaneous combustion of coal. This number is expected to rise with greater coal consumption, deeper mines, and increased utilization of lower rank coals. "Hot" coal is also a problem often encountered in the transport and storage of coal. In addition, the self-heating of sulfide minerals has caused fires in noncoal mines. To address these concerns, the U.S. Bureau of Mines has an ongoing spontaneous combustion research program. This paper highlights several aspects of the research. These include (1) laboratory studies to evaluate the self-heating tendencies of coals and sulfide minerals and to assess the factors that affect the self-heating process, such as the heat of wetting, (2) the development of a sealed-flask test, based on the absorption of oxygen, for determining the self-heating potential of coals in the field, (3) studies to assess the efficiency of additives and sealants for suppressing spontaneous combustion, (4) large-scale experiments to examine the self-heating of a 13-ton coalbed, and (5) the development of a formula for predicting the spontaneous combustion potential of bituminous coals and its utilization by a mining company to evaluate the self-heating hazards in several mines.